Generative adversarial network based on chaotic time series
This work addresses data and randomness requirements in GANs for image generation, but it is incremental as it modifies the latent variable source rather than introducing a new paradigm.
The authors tackled the need for large amounts of genuine training data and pseudorandom numbers in GANs by using chaotic time series from semiconductor lasers as latent variables, resulting in enhanced similarity in proximity for generated images without severe degradation in versatility and elimination of negative autocorrelation signatures.
Generative adversarial network (GAN) is gaining increased importance in artificially constructing natural images and related functionalities wherein two networks called generator and discriminator are evolving through adversarial mechanisms. Using deep convolutional neural networks and related techniques, high-resolution, highly realistic scenes, human faces, among others have been generated. While GAN in general needs a large amount of genuine training data sets, it is noteworthy that vast amounts of pseudorandom numbers are required. Here we utilize chaotic time series generated experimentally by semiconductor lasers for the latent variables of GAN whereby the inherent nature of chaos can be reflected or transformed into the generated output data. We show that the similarity in proximity, which is a degree of robustness of the generated images with respects to a minute change in the input latent variables, is enhanced while the versatility as a whole is not severely degraded. Furthermore, we demonstrate that the surrogate chaos time series eliminates the signature of generated images that is originally observed corresponding to the negative autocorrelation inherent in the chaos sequence. We also discuss the impact of utilizing chaotic time series in retrieving images from the trained generator.